15. THERMOBAROMETRIC CALCULATION OF PEAK METAMORPHIC CONDITIONS OF THE SULLIVAN DEPOSIT Glen R. De Paoli1 and David R.M. Pattison2

1 2739 4th Ave N.W., Calgary, Alberta T2N 0P9 2 Department of Geology and Geophysics, The University of Calgary, 2500 University Drive N.W., Calgary, Alberta T2N 1N4

ABSTRACT The Sullivan Zn-Pb-Ag mine in southeastern British Columbia is a mid-Proterozoic sedimentary exhalative deposit that has been metamorphosed to transitional greenschist-amphibolite grade. The mineral textures are predominantly of metamorphic origin. There are no significant differences in silicate mineral compositions between sulphide-rich litholo- gies and silicate-rich lithologies. High modal abundance of Mn-rich garnet in some rocks indicates a Mn-rich phase was likely part of the pre-metamorphic mineral assemblage. Metamorphic conditions were estimated using multi-equi- librium thermobarometric techniques involving silicate-carbonate equilibria. Peak metamorphic temperature estimated by the garnet-biotite Fe-Mg exchange equilibrium is 450oC ± 50oC. Lower temperature estimates from some samples are interpreted to record the temperature of cessation of garnet growth prior to the attainment of peak metamorphic tem- perature. Peak metamorphic pressure determined from equilibria applicable to the assemblage garnet-biotite-mus- covite-chlorite-calcite-quartz-fluid is 380 MPa ± 100 MPa (equivalent to approximately 14 km of burial). The fluid composition accompanying this pressure estimate is XH2O = 0.38, XCO2 = 0.62 ± 0.07. The P-T estimates are con- sistent with burial metamorphism during Proterozoic times. INTRODUCTION PETROGRAPHY The Sullivan mine, in southeastern British Columbia, is a Samples were classified into three rock types on the basis of sedimentary exhalative Zn-Pb-Ag orebody which was bulk mineralogy: silicate matrix rocks, sulphide matrix deposited in mid-Proterozoic times. Regional metamorphic rocks, and amphibolite. None of the samples used for P- grade in the vicinity of the orebody is greenschist facies T-X estimates shows any textural or compositional evidence (biotite-chlorite), but amphibolite grade areas containing of retrograde alteration. Localized retrograde alteration is garnet and sillimanite occur in the Mathew Creek area found in the southeast section of the orebody (Ethier et al., approximately 10 km to the southwest (Leech, 1962; 1976; Campbell and Ethier, 1983). Mineralogical evidence McFarlane and Pattison, 2000; Fig. 15-1). The deposit itself includes: chloritized biotite; biotite with compositional zon- is anomalous with respect to the immediately surrounding ing; stilpnomelane overgrowths on sulphides and silicates; rocks in that it contains abundant metamorphic garnet. hematite staining; and poikilitic magnetite grains enclosing Many characteristics of the deposit have been affected by pyrite. the metamorphism, including fluid inclusions, stable iso- topes, and the textures and mineralogy of the ore, gangue, Silicate matrix rocks and pre-metamorphic hydrothermal alteration zones. An Silicate matrix rocks are represented by the ‘waste beds’ important step in understanding the effects of a metamorphic (sulphide poor siliciclastic interbeds) within the bedded ore overprint is a well-constrained estimate of metamorphic and contain less than 10% sulphides. The overall mineral temperature, pressure, and fluid composition (P-T-X). assemblage is Qtz + Ms ± Bt ± Chl ± Grt ± Pl ± Cal, with Po Previous estimates of metamorphic conditions for the ± Py ± Sph ± Gn. Sullivan deposit are summarized in Table 15-1. Until this study, no attempt has been made to estimate metamorphic Porphyroblasts conditions using the techniques of multi-equilibrium ther- Porphyroblasts in silicate matrix rocks include garnet, mobarometry applied to the silicate-carbonate minerals in biotite, chlorite, and calcite. In some samples, distinguish- the orebody. An estimate by these methods has the advan- ing between porphyroblasts and matrix was not feasible due tage of determining temperature, pressure, and fluid compo- to uniformity of grain size. Garnet is found in most samples sition simultaneously. Some of the information in this con- and its abundance and grain size is approximately propor- tribution is presented in detail in De Paoli and Pattison tional to sulphide abundance. Garnets range from 0.1 to 1.0 (1995). Mineral abbreviations are given in Table 15-2. mm in diameter and occur as isolated idiomorphic grains Sample locations are given in Figure 15-2. with inclusion-rich cores, and as aggregates of poikiloblastic grains rimming pyrrhotite laths (Fig. 15-3a). Garnet inclu- sions include quartz, calcite, chlorite, and pyrrhotite. Biotite and chlorite porphyroblasts occur as randomly oriented idiomorphic grains up to 1.0 mm across. Calcite is rare

De Paoli, G.R. and Pattison, D.R.M. 2000: Thermobarometric Calculation of Peak Metamorphic Conditions of the Sullivan Deposit; in The Geological Environment of the Sullivan Deposit, British Columbia, (ed. J.W. Lydon, J.F. Slack, T. Höy and M.E. Knapp; Geological Association of Canada, Mineral Deposits Division, MDD Specal Volume No. 1, p. G.R. De Paoli and D.R.M. Pattison

......

......

Figure 15-2. Outline of the Sullivan orebody showing sample loca- tions. Samples DS-47, DS-85, and UCG-84 are from bedded ore on the eastern margin of the deposit, but their exact locations are uncertain.

± Py ± Grt ± Bt ± Ms ± Qtz ± Chl ± Cal. Differentiation of matrix vs. porphyroblasts was not always possible due to complex sulphide mineral textures and the broad range in size of both the sulphide and silicate minerals. Sulphide Minerals The major sulphide minerals include pyrrhotite, spha- Figure 15-1. Simplified geological map of the Sullivan Mine Region lerite, galena (± boulangerite), and minor pyrite. The sul- (Leech, 1957, 1962; Reesor, 1973; Höy, 1993; Höy et al, 1995; Read et al. 1991). phide minerals show the following varieties: anhedral equigranular pyrrhotite, galena, and sphalerite; massive among the silicate matrix samples examined (2 out of 9 sam- pyrrhotite or sphalerite containing small anhedral blebs of ples, with 9 and 3 mode % calcite respectively). In places sphalerite, pyrrhotite, or galena; and a crudely banded tex- where it was found, it ranges from matrix size to 0.5 mm, ture of anhedral pyrrhotite, sphalerite, and galena. Pyrite with the larger grains commonly rimming pyrrhotite laths. usually exhibits a euhedral, porphyroblastic texture. Matrix Silicate and Carbonate Minerals The matrix is composed of roughly equal amounts of fine- Silicate and carbonate minerals in sulphide matrix rocks grained (10 to 50 µm) quartz and muscovite. In some rocks, include quartz, muscovite, garnet, biotite, and chlorite, and biotite and chlorite are fine-grained enough to be considered carbonate minerals with varying amounts of Ca, Fe, Mg, and matrix rather than porphyroblasts. Many quartz grains show Mn (see Mineral Chemistry section below). Silicate and undulose extinction and annealed textures. Plagioclase, pos- carbonate minerals tend to be coarser-grained and less euhe- sibly peristerite, (see Mineral Chemistry section below) was dral than in silicate matrix rocks. Carbonate minerals are identified in the matrix of one sample with the electron more abundant than in silicate matrix rocks (2 of three sam- microprobe. Accessory minerals include sulphides (mostly ples with 10 mode % carbonate in each compared to values pyrrhotite laths oriented parallel to bedding), titanite, apatite, given above for silicate matrix rocks). Quartz and carbonate and zircon. are generally anhedral. Larger and more euhedral garnet, biotite, muscovite, and chlorite porphyroblasts reach 2 mm Sulphide Matrix Rocks in size. Garnet-rich laminae are commonly found at the con- tact between sulphide-rich and silicate-rich bands (Fig. 15- Sulphide matrix rocks include samples from the sulphide- 3b). rich layers in the bedded ore and contain > 40% sulphide minerals. The overall mineral assemblage is Po ± Sph ± Gn 216 Thermobarometric Calculation of Peak Metamorphic Conditions

Table 15-1. Previous estimates of metamorphic conditions The virtual absence of analyses between an3 and an13 sug- for the Sullivan Orebody. gests the presence of peristerite. Peristerite is plagioclase in method temperature pressure reference o which albitic and more intermediate (more Ca-rich) plagio- ( C) (MPa) clase coexist, with exact compositions controlled by a pres- oxygen isotope fractionation 400 - 560 Ethier et al. (1976) o between quartz and magnetite sure-temperature sensitive solvus. However, at 450 C and sulphur isotope fractionation 300 - 340 Ethier et al. (1976) 380 MPa (the preferred P-T estimate, see P-T-X Estimate between sphalerite and galena section below) the peristerite gap spans approximately an3 to arsenopyrite geothermometer 400 - 490 Ethier et al. (1976) an20, not an3 to an13 (Crawford, 1966; Maruyama et al. (Kretschmar and Scott, 1976) 1982). The possibility exists that the plagioclase composi- sphalerite geobarometer 500 Ethier et al. (1976) (Scott, 1973) tions may represent not only metamorphic peristerite, but also fluid inclusions in quartz 200 - 415 Leitch (1992) relict igneous compositions and albite, the latter related to sphalerite geobarometer 375 450 Lydon and Reardon pre-metamorphic hydrothermal alteration. Interpretation of (Toulmin et al. 1991) (2000) the significance of the compositional gap with respect to metamorphic temperature and pressure requires reliable inter- Amphibolite pretation of peristerite textures. Efforts to image the textures of individual feldspar grains with a scanning electron micro- The amphibolite represents metamorphosed gabbro and dior- scope were unsuccessful due to small grain size and the very ite rocks of the Moyie sills which intrude the western edge of slight contrast in mean atomic number between minerals so the deposit. Metamorphic minerals and textures in this sam- close in chemical composition. Similar attempts to determine ple have overprinted most of its original igneous character. compositional/textural relationships for plagioclase in the The rock is composed of equigranular, subhedral to euhedral amphibolite were also unsuccessful, with compositions rang- amphibole, chlorite, quartz, plagioclase, clinozoisite, and cal- ing from pure albite to an with no compositional gap. see 17 cite. Plagioclase shows complex textures ( Mineral Consequently, we feel that we do not sufficiently understand below Chemistry section ) and is strongly sericitized. the plagioclase relations to use them for P-T interpretations. Pyrrhotite and ilmenite are disseminated throughout. Ilmenite is nearly always rimmed by titanite and/or clinozoisite. Carbonate analyses differ noticeably between silicate matrix and sulphide matrix rocks. The composition of car- MINERAL CHEMISTRY bonate in silicate matrix rocks is near pure calcite, with less than 2 wt.% combined FeO, MgO, and MnO. Carbonate in Minerals were analyzed by electron microprobe at the sulphide matrix rocks have significant siderite (up to 14.21 University of Calgary. Analytical methods are described in wt.% FeO), dolomite (up to 15.98 wt.% MgO), and De Paoli and Pattison (1995). Full mineralogical composi- rhodocrosite (up to 4.58 wt.% MnO) components. tions are given in Appendix 12. In many metamorphosed sulphide ore deposits, particu- P-T-X Estimate larly those of higher metamorphic grade, silicate minerals Metamorphic pressure, temperature and fluid composition such as garnet, biotite and chlorite tend to be more Mg-rich estimates were made using the multi-equilibrium thermo- and less Fe-rich in sulphide-rich rocks than in sulphide-poor rocks (Froese, 1969; Bachinski, 1976; Hutcheon, 1979; Nesbitt, 1982). Within the Sullivan deposit there are differ- ences in mineral composition between silicate-rich and sul- Table 15-2. Mineral, end members, and abbreviations (after phide-rich rocks, but two distinct populations of Fe/(Fe+Mg) Kretz, 1983). ratios cannot be discerned. mineral symbol end member symbol As an example, figure 15-4 shows Fe/(Fe+Mg) vs. Xsps garnet Grt almandine (Fe) alm in garnet rims for both the silicate and sulphide lithologies pyrope (Mg) prp (analyses in Table 15-3). Although the range of Fe/(Fe+Mg) spessartine (Mn) sps ratios for garnets in sulphide-rich rocks is slightly lower grossular (Ca) grs (from 0.888 to 0.953) than in silicate rocks (from 0.904 to biotite Bt annite (Fe) ann 0.963), there is a large overlap in the range of values between phlogopite (Mg) phl the two rock types. There is similar overlap in the variation of manganese content in garnet, with Xsps in sulphide-rich plagioclase Pl anorthite (Ca) an rock ranging from 0.277 to 0.518, and Xsps in silicate-rich albite (Na) alb rocks ranging from 0.090 to 0.454. Biotite compositions orthoclase (K) or show a similar trend in Fe/(Fe+Mg) as garnet. The range of chlorite Chl clinochore (Mg) cln Fe/(Fe+Mg) in biotite in sulphide-rich rocks (from 0.315 to daphnite (Fe) dph 0.596) is lower than in silicate-rich rocks (from 0.396 to muscovite Ms 0.687), but the range of overlap is large. calcite Cal Electron microprobe examination identified fine-grained quartz Qtz (50 µm) plagioclase in the matrix of one silicate matrix sam- pyrrhotite Po ple. Compositions within a single grain and between grains range from pure albite to an (total of 78 analyses), with a pyrite Py 28 sphalerite Sph compositional gap between an3 and an13 (one value of an7). galena Gn 217 G.R. De Paoli and D.R.M. Pattison

Table 15-3. Selected garnet-biotite compositional data and temperatures at 380 MPa from the Berman (1990) calibration of the garnet-biotite Fe-Mg exchange geothermometer. Data for rim analyses unless otherwise indicated. Garnetb Biotitec sample lithology XFe XMg XMn XCa Fe/ XFe XMg XAlVI XTi Fe/ mode % T (oC) a Fe+Mg Fe+Mg Chld S1-57 silicate 0.359 0.038 0.454 0.138 0.904 0.334 0.514 0.103 0.033 0.394 trace 377 S3-11 silicate 0.726 0.031 0.143 0.100 0.959 0.598 0.267 0.095 0.032 0.691 5 460 (core) 0.678 0.028 0.198 0.096 0.960 448 S3-14 sulphide 0.338 0.038 0.518 0.106 0.899 0.284 0.617 0.061 0.022 0.315 ---- 296 S3-16 silicate 0.723 0.028 0.219 0.024 0.963 0.590 0.256 0.112 0.025 0.697 ---- 411 S3-18 sulphide 0.666 0.033 0.277 0.017 0.953 0.518 0.351 0.080 0.041 0.596 15 364 (core) 0.631 0.024 0.285 0.051 0.963 330 S3-22 silicate 0.609 0.031 0.320 0.037 0.952 0.497 0.356 0.107 0.027 0.583 ---- 365 S7-150 silicate 0.566 0.043 0.337 0.040 0.929 0.504 0.350 0.092 0.039 0.590 20 451 S7-153 silicate 0.610 0.055 0.294 0.034 0.917 0.464 0.393 0.098 0.031 0.541 5 448 S7-155 silicate 0.604 0.050 0.277 0.057 0.924 0.441 0.413 0.102 0.028 0.516 trace 416 (core) 0.568 0.049 0.273 0.098 0.921 439 DS-47 silicate 0.752 0.036 0.123 0.084 0.954 0.551 0.303 0.105 0.036 0.645 40 431 DS-85 silicate 0.783 0.036 0.090 0.084 0.956 0.572 0.273 0.117 0.032 0.677 40 456 (core) 0.509 0.018 0.228 0.172 0.966 403 UCG-84 sulphide 0.427 0.054 0.476 0.036 0.888 0.341 0.531 0.092 0.021 0.391 ---- 385 (core) 0.304 0.031 0.563 0.093 0.907 358 aindicates silicate matrix or sulphide matrix rock bXi = i/(Fe2+ + Mg + Mn + Ca) cXi = i/(Fe + Mg + Mn + AlVI + Ti) dby visual estimation

barometric computer program TWEEQU, version 1.02 such as rhodochrosite (MnCO3), or pyrolusite (MnO2). (Berman, 1988, 1991). A description of the method is pro- Pyrolusite, or other oxides, are unlikely to have been present vided in Berman (1991). Most samples only contained suf- in the deposit prior to metamorphism because of the reduc- ficient mineral phases to obtain a temperature estimate (gar- ing conditions of the sediments (W. Goodfellow, pers. net-biotite Fe-Mg exchange). One sample, containing the comm., 1995). Reactions involving rhodochrosite and detri- mineral assemblage Grt-Bt-Ms-Chl-Cal-Qtz, allowed a com- tal clay minerals (represented by kaolinite below) may have plete P-T-X estimate. produced Mn-rich garnet by a reaction of the type: (1) 3MnCO + Al Si O (OH) + SiO = Geothermometry 3 2 2 5 4 2 Mn3Al2Si3O12 + 3CO2 + 2H2O Table 15-3 shows selected garnet and biotite compositions (rhodochrosite) + (kaolinite) + (quartz) = and temperature estimates (assuming a pressure of 380 MPa) (spessartine) + (CO2) + (H2O) for 12 samples. Temperature estimates range from 296 to o Relatively manganese-rich carbonate (4.58 wt.%) was iden- 460 C. De Paoli and Pattison (1995) argued that the most tified in a sulphide matrix rock (see Mineral Chemistry sec- likely explanation for the range in temperature estimates was tion above), further suggesting that Mn-carbonates may have that garnet stopped growing prior to the attainment of peak been a significant part of the original depositional mineralo- metamorphic temperature in those samples giving lower gy of the deposit. temperature estimates. Petrographic evidence indicates chlorite may also have Some garnets may have stopped growing earlier than oth- been an important reactant phase during garnet growth. ers due to the loss of a reactant phase involved in garnet Those samples which give low temperature estimates (< growth - in particular, the phase or phases supplying man- 430oC) contain little or no chlorite, whereas samples giving ganese. Garnet has been shown to be stable at lower tem- higher temperature estimates (> 430oC, including all five peratures in a geochemical system containing manganese samples selected for the temperature estimate, see below) than in a pure Fe-Mg system (Symmes and Ferry, 1992). contain significant chlorite (5 - 40 mode %). Garnet growth Figure 15-1 shows garnet is not part of the regional meta- may have occurred via a chlorite-consuming reaction of the morphic assemblage in the immediate vicinity of the type: Sullivan orebody. Consequently, the most likely explanation for the presence of garnet within the orebody is anomalous- (2) Chlorite(Mn) + muscovite + quartz = ly Mn-rich bulk rock compositions within the deposit com- garnet (Mn) + biotite + H2O pared to the Aldridge Formation in general. and ceased at the temperature at which the chlorite was Several reactant phases may have supplied manganese to exhausted. In those rocks containing a low modal abun- the garnets in the deposit. The high manganese content and dance of relatively Mn-poor garnet, reaction (2) alone may high modal abundance of garnet in many samples, especial- have produced the garnets, whereas in rocks with higher ly sulphide-matrix samples (e.g. 37 wt. % MnO and 85 mode modal abundance of relatively Mn-rich garnet, reaction (2) % garnet in one sample), requires a Mn-rich reactant phase, may have proceeded either simultaneously with, or follow-

218 Thermobarometric Calculation of Peak Metamorphic Conditions a b

Figure 15-3. Photomicrographs of typical mineral textures in the Sullivan orebody. (a) Sample S3-11 from a waste band within the bedded ore showing garnet (Grt) rimming pyrrhotite (Po). (b) Sample UCG-84, garnets (Grt) concentrated along contact between silicate waste layer and sulphide layer within bedded ore. ing, the completion of reaction (1). Sphalerite and pyrrhotite The average pressure estimate of the three locations in may also have been subordinate sources of manganese. sample S3-11 is 380 MPa ± 100 MPa (Table 15-4). The A temperature estimate for the deposit of 450oC ± 50oC uncertainty results from recalculating the equilibrium at two was made from the average of the five highest estimates, standard deviations of the mineral analyses. samples S3-11, S7-150, S7-153, DS-47, and DS-85. The quoted uncertainty results from recalculating the equilibria Fluid Composition at two standard deviations of the mineral analyses (Ferry and The average of the fluid compositions from the three loca- Spear, 1978). tions in sample S3-11 is XH2O = 0.38, XCO2 = 0.62 (Table 15-4). The deviation of the maximum XH2O (0.44) and the Geobarometry minimum XH2O (0.31) from the mean XH2O (0.38) Among the many different rock types examined from the between the three determinations, a value of ± 0.07, is taken Sullivan deposit, only one sample with a mineral assemblage as an indication of the uncertainty in the fluid composition suitable for geobarometry was identified. This is sample S3- estimate. This relatively CO2-rich fluid estimate is particu- 11 from a silicate matrix rock with the mineral assemblage lar to this sample, and is not necessarily representative of Grt-Bt-Ms-Chl-Cal-Qtz. Equivocal feldspar textures and metamorphic fluid compositions for the whole deposit. compositions (see Mineral Chemistry section above) did not The pressure estimate was calculated assuming the meta- allow application of pressure-sensitive, fluid composition- morphic fluid was an H2O-CO2 binary. This assumption is independent equilibria involving plagioclase end members. reasonable, as the concentration of other possible fluid phas- In sample S3-11, three locations were selected where all the es such as CO, CH4, H2, H2S, SO2 and COS determined minerals of interest were in close proximity. Each location using the computer software and fluid property data of was treated as a separate chemical system, and the P-T- Connolly and Cesare (1993) is negligible (Table 15-5). The XCO2 determination for each location was averaged to values in Table 15-5 were generated assuming the presence determine the P-T-XCO2 estimate for the sample. This of graphite and a pyrite-pyrrhotite sulphur activity buffer so approach accounts for small scale variations in bulk compo- as to give the maximum possible concentrations of the trace sition within the sample. Pressure-temperature, pressure- fluid phases. C-O-H-S trace species concentrations in the XCO2, and temperature-XCO2 diagrams for location #2 in metamorphic fluid do not reach significant concentrations at sample S3-11 are given in Figures 15-6a,b and c respective- this metamorphic grade regardless of the H2O/CO2 ratio, and ly. A summary of the P-T-X results for all three locations in therefore metamorphic fluids for the deposit as a whole were sample S3-11 are given in Table 15-4. likely an H2O/CO2 binary. With the exception of the garnet-biotite Fe-Mg equilibri- While an XCO2 of 0.62 may seem high for greenschist um (curve #1 in Figures 15-6a,b,c), equilibria from the min- grade rocks, carbonate is a significant component of the bulk eral assemblage Grt-Bt-Ms-Chl-Cal-Qtz involve a fluid mineralogy of the Sullivan orebody (Hamilton et al. 1982, phase. The position of the curves for these equilibria on P- T-X diagrams changes as the H2O/CO2 ratio of the fluid is Table 15-4. Temperature, pressure and fluid composition varied. Without an independent determination of the meta- estimates. Sample S3-11. morphic fluid composition, the H O/CO ratio which gave 2 2 o the tightest intersection of the equilibrium curves was select- location temperature ( C) pressure (MPa) XH2O ed. The high angle of intersection of the equilibrium curves 1 470 480 0.44 in Figure 15-6b demonstrates that the equilibrium fluid com- 2 460 360 0.38 position is tightly constrained, as even a small variation of 3 450 300 0.31 the H2O/CO2 ratio destroys the tight intersection. average 460 380 0.38 219 G.R. De Paoli and D.R.M. Pattison

0.98 500

silcate matrix silcate matrix sulphide matrix sulphide matrix 0.96 450

0.94 400

0.92 Fe/(Fe+Mg) 350 T(C) at 380 MPa

0.90 300

0.88 250 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Xsps Xsps

Figure 15-4. Fe/(Fe+Mg) vs. Xsps in garnet. Rim analyses plotted. Figure 15-5. Temperature estimate (oC) at 380 MPa vs. Xsps in gar- Open circles denote silicate matrix samples, filled circles denote sul- net, using rim analyses. Temperature estimate from the Berman phide matrix samples. While garnets in a sulphide matrix tend to (1990) calibration of the garnet-biotite Fe-Mg exchange geother- have lower Fe/(Fe+Mg) and higher Xsps than garnets in a silicate mometer. See Table 15-1 for garnet and biotite compositional data. matrix, the broad range of compositions from both rock types over- Figure shows lower temperature estimates come from samples with lap significantly. Garnets with higher Xsps have lower Fe/(Fe+Mg) in higher Xsps in garnet. both rock types.

Leitch, 1992), and may have been more abundant in the pre- Taking into consideration the maximum limit of the pres- metamorphic mineral assemblage. Significant CO2 will sure estimate (380 MPa+ 100 MPa = 480 MPa), a pressure have been introduced into the metamorphic fluid via reaction of 480 MPa is equivalent to 18 km of burial and gives a geot- (1) shown above. hermal gradient of only 25oC/km. Eighteen km of burial is close to an estimate of 17 km of burial during Jurassic- DISCUSSION Cretaceous times for the Mount Fisher area, 35 km east of There is considerable uncertainty as to the age of metamor- the Sullivan deposit (McMechan and Price, 1982). While phism of the Sullivan deposit. The Sullivan region has a long there is evidence for Mesozoic deformation and retrograde and complex tectonic and thermal history, and several meta- metamorphism of the Sullivan orebody, both are interpreted morphic episodes have been inferred: ca. 1330 Ma (Ryan to post-date maximum metamorphic conditions (McClay, and Blenkinsop, 1971; Ross et al., 1992; Schandl et al., 1983; Lydon and Reardon, 2000). 1993; Anderson and Parrish, 2000; McFarlane and Pattison, The minimum limit of the pressure estimate (380 MPa - 2000; Schandl and Davis, 2000); ca. 1050 - 1100 Ma 100 MPa = 280 MPa) gives a geothermal gradient of (Anderson and Parrish, 2000); and Jurassic-Cretaceous time 45oC/km, which would almost certainly require a thermal (McMechan and Price, 1982; Lydon and Reardon, 2000). anomaly. Causes for a thermal anomaly could include mag- The age of the peak metamorphic event at Sullivan (and pos- matic intrusions (e.g. the nearby Hellroaring Creek stock) or sible subsequent ones) would best be determined by a pres- heat associated with a possible Proterozoic orogenic event sure, temperature, and fluid composition estimate made from (McFarlane and Pattison, 2000). a mineral assemblage which includes a metamorphic miner- al from which an age can be determined, such as titanite, but to date this has not been accomplished.

The metamorphic pressure estimate of 380 MPa is equiv- Table 15-5. Fluid species concentrations for sample S3-11. alent to approximately 14 km of lithostatic load (assuming XH2O and XCO2 from TWEEQU software (Berman, 1988, an average rock density of 2.7 g/cm3). For 450oC, this gives 1991), others from Connolly and Cesare (1993). Not nor- a geothermal gradient of 32oC/km, similar to the average malized to 100%. o geothermal gradient today of 30 C/km (Best, 1982), and species mole fraction consequently, consistent with burial metamorphism. With respect to the timing of metamorphism, 14 km of burial is in H2O 0.32 accord with the estimated thickness of Purcell Supergroup CO2 0.68 CO 1.25 * 10-4 sediments overlying the deposit (Höy et al, 1995, Höy, pers. -4 CH4 3.07 * 10 comm., 1997). However, there remains the question of -5 unknown thicknesses of sediment removed at numerous H2 5.74 * 10 -3 unconformities within the Purcell Supergroup and younger H2S 1.31 * 10 -10 strata. SO2 6.22 * 10 COS 3.30 * 10-5 220 Thermobarometric Calculation of Peak Metamorphic Conditions

Figure 15-6. (a) Pressure-temperature, (b) Pressure-XCO2, and (c) Temperature-XCO2 diagrams for location #2, sample S3-11. Mineral assemblage is: Grt-Bt-Ms-Chl-Cal-Qtz. Three independent equilibria. Intersection temperature = 460oC; pressure = 360 MPa; XH2O = 0.38. Equilibria as follows: 1. alm + phl = prp + ann 2. 4 ann + 3 cln + ms + 3 Qtz = 4 alm + 5 phl + 12 H2O 3. alm + 3 cln + ms + 3 Qtz = 5 prp + ann + 12 H2O 4. 3 cln + ms + 3 Qtz = 4 prp + phl +12 H2O 5. alm + ms + 6 Cal + 3 Qtz = 2 grs + ann + 6 CO2 6. prp + ms + 6 Cal + 3 Qtz = 2 grs + phl + 6 CO2 7. 2 grs + 5 ann + 3 Cln + 6 CO2 = 5 alm = 5 phl + 6 Cal + 12 H2O 8. 2 grs + cln + 6 CO2 = 5 prp + 6 Cal + 12 H2O 9. 3 cln + 5 ms + 24 Cal + 15 Qtz = 8 grs + 5 phl + 24 CO2 + 12 H2O The left side of each equilibrium corresponds to the low temperature side of the curves in Fig. 15-6a and 15-6c. For Fig. 15-6b the topolo- gies can be derived from Fig. 15-6a and 15-6c. From De Paoli and Pattison (1995).

CONCLUSIONS rocks in the deposit. Metamorphic fluids were essentially a H O-CO binary. The pressure-temperature estimate is con- The Sullivan orebody has been metamorphosed to transi- 2 2 sistent with burial metamorphism, most likely of Proterozoic tional greenschist-amphibolite facies (chlorite-biotite-gar- age. net). Mineral textures in the deposit are primarily of meta- morphic origin. The abundance of garnet within the deposit ACKNOWLEDGMENTS compared to the regional chlorite-biotite assemblage is like- ly due to more Mn-rich bulk compositions within the ore- Funding for this research was provided by Natural Sciences body. The high modal abundance of Mn-rich garnet in some and Engineering Research Council of Canada grant 037233 rocks indicates a Mn-rich phase was likely part of the pre- to David R.M. Pattison. The authors wish to thank J. W. metamorphic mineral assemblage. Peak metamorphic P-T- Lydon, T. Höy, and C.H.B. Leitch for their constructive X conditions were estimated using TWEEQU computer soft- reviews of this contribution. The authors also benefited ware (Berman, 1991). Peak metamorphic temperature from from numerous discussions with other contributors to this garnet-biotite Fe-Mg exchange equilibrium is 450oC ± 50oC. volume. Lower temperature estimates from some samples are inter- preted to record the temperature of cessation of garnet REFERENCES growth prior to peak metamorphic conditions. Peak meta- Anderson, H.E. and Parrish, R.R. morphic pressure determined from the assemblage garnet- 2000: U-Pb geochronological evidence for the geological history biotite-muscovite-chlorite-calcite-quartz is 380 MPa ± 100 of the Belt-Purcell Supergroup, southeastern B.C.; in The MPa (equivalent to 14 km of burial). The fluid composition Geological Environment of the Sullivan Deposit, British Columbia, (ed.) J.W. Lydon, J.F. Slack, T. Höy, and M.E. required for this pressure estimate is XH2O = 0.38, XCO2 = 0.62, ± 0.07. This fluid composition is specific to one sam- Knapp; Geological Association of Canada, Mineral Deposits Division, MDD Special Volume No. 1, p. ple and may not be representative of fluid compositions in all

221 G.R. De Paoli and D.R.M. Pattison

Bachinski, J.R. Kretschmar, U. and Scott, S.D. 1976: Metamorphism of cuperiferous iron deposits, Notre Dame 1976: Phase relations involving arsenopyrite in the system Fe-As- Bay, Newfoundland; Economic Geology, v. 71, p. 443-452. S and their applications; Canadian Mineralogist, v. 14, Berman, R.G. p. 364-386. 1988: Internally-consistent thermodynamic data for stoichiomet- Kretz, R. ric minerals in the system Na2O-K2O-CaO-FeO-Fe2O3- 1983: Symbols for rock-forming minerals; American TiO2-H2O-CO2; Journal of Petrography, v. 75, p. 328-344. Mineralogist, v. 68, p. 277-279. 1990: Mixing properties of Ca-Mg-Fe-Mn garnets; The American Leech, G.B. Mineralogist, v. 75, p. 328-344. 1957: St. Mary Lake, Kootenay District, British Columbia 1991: Thermobarometry using multiequilibrium calculations: a (82F/9); Geological Survey of Canada, Map 15-1957, scale new technique with petrological applications; Canadian 1:63 360. Mineralogist, v. 29, p. 833-855. 1962: Metamorphism and Granitic intrusions of Precambrian age Best, M.G. in southeastern British Columbia; Geological Survey of 1982: Igneous and metamorphic Petrology; W.H. Freeman and Canada, Paper 62-13, 8 p. Company, New York, 690 p. Leitch, C.H.B. Campbell, F.A. and Ethier, V.G. 1992: Mineral chemistry of selected silicates, carbonates, and sul- 1983: Environment of deposition of the Sullivan orebody; phides in the Sullivan and North Star stratiform Zn-Pb Mineralium Deposita, v. 18, p. 39-55. deposits, British Columbia, and in district-scale altered and Crawford, M.L. unaltered sediments; in Current Research, Part E; 1966: Composition of plagioclase and assorted minerals in some Geological Survey of Canada Report 92-1E, p. 83-93. schists from Vermont, USA, and south Westland, New Lydon, J.W. and Reardon, N.C. Zealand, with inferences about the peristerite solvus; 2000: Sphalerite compositions of the Sullivan deposit and their Contributions to Mineralogy and Petrology, v. 13, p. 269-294. implications for its metamorphic history; in The Geological Connolly, J.A.D. and Cesare, B. Environment of the Sullivan Deposit, British Columbia, 1993: C-O-H-S fluid compositions and oxygen fugasity in (ed.) J.W. Lydon, J.F. Slack, T. Höy, and M.E. Knapp; graphitic metapelites; Journal of Metamorphic Geology, Geological Association of Canada, Mineral Deposits v. 11, p. 379-388. Division, MDD Special Volume No. 1, p. De Paoli, G.R. and Pattison, D.R.M. Maruyama, S., Liou, L.G., and Suzuki, K. 1995: Constraints on temperature-pressure conditions and fluid 1982: The peristerite gap in low-grade metamorphic rocks; composition during metamorphism of the Sullivan orebody, Contributions to Mineralogy and Petrology, v. 81, p. 268-276. Kimberley, British Columbia, from silicate-carbonate equilib- McClay, K.R. ria; Canadian Journal of Earth Sciences, v. 32, p. 1937-1949. 1983: Structural evolution of the Sullivan Fe-Pb-Zn-Ag orebody, Ethier, V.G., Campbell, F.A., Both, T.A., and Krouse, H.R. Kimberley, British Columbia, Canada; Economic Geology, 1976: Geological setting of the Sullivan orebody and estimates of v. 78, p. 1398-1424. temperatures and pressure of metamorphism; Economic McFarlane, C.R.M. and Pattison, D.R.M. Geology, v. 71, p. 1570-1588. 2000: Geology of the Mathew Creek metamorphic zone, south- Ferry, J.M. and Spear, F.S. east British Columbia: a window into Middle Proterozoic 1978: Experimental calibration of the partitioning of Fe and Mg metamorphism in the Purcell Basin; Canadian Journal of between biotite and garnet; Contributions to Mineralogy Earth Sciences, v. 37, p. 1073-1092. and Petrology, v. 66, p. 113-117. McMechan, M.E. and Price, R.A. Froese, E. 1982: Superimposed low-grade metamorphism in the Mount 1969: Metamorphic rocks from the Coronation mine and sur- Fisher area, southeastern British Columbia - implications rounding area; Canadian Geological Survey, Paper 68-5, for the East Kootenay orogeny; Canadian Journal of Earth p. 55-77. Sciences, v. 19, p. 476-489. Hamilton, J.M., Bishop, D.T., Morris, H.C., and Owens, O.E. Nesbitt, B.E. 1982: Geology of the Sullivan orebody, Kimberley, British 1982: Metamorphic sulphide-silicate equilibrium in the massive Columbia, Canada; in Precambrian Sulphide Deposits, sulphide deposits at Ducktown, Tennessee; Economic (ed.) R.W. Hutchinson, C.D. Spence and J.M. Franklin; Geology, v. 77, p. 364-378. Geological Association of Canada, Special Paper 25, H.S. Read, P.B., Woodsworth, G.J., Greenwood, H.J., Ghent, E.D., Robinson Memorial Volume, p. 598-665. and Evenchick, C.A. Höy, T. 1991: Metamorphic map of the Canadian Cordillera; Geological 1993: Geology of the Purcell Supergroup in the Fernie West-half Survey of Canada, Map 1714A. map-area, southeastern British Columbia; British Columbia Reesor, J.E. Ministry of Energy, Mines and Petroleum Resources, 1973: Geology of the Lardeau map-area, east-half, British Bulletin 84, 157 p. Columbia; Geological Survey of Canada, Memoir 369, 129 p. Höy, T., Price, R.A., Legun, A., Grant, B., and Brown, D.A. Ross, G.M., Parrish, R.R., and Winston, D. 1995: Purcell Supergroup, southeastern British Columbia, compi- 1992: Provenance and Pb-U geochronology of the lation map; British Columbia Ministry of Energy, Mines, Mesoproterozoic Belt Supergroup (northwestern United and Petroleum Resources, Geoscience Map 1995-1, scale States): implications for the age of deposition and pre- 1:250 000. Panthalassa plate reconstructions; Earth and Planetary Hutcheon, I. Science Letters, v. 113, p. 57-76. 1979: Sulphide-oxide-silicate equilibria; Snow Lake, Manitoba; Ryan, B.D. and Blenkinsop, J. American Journal of Science, v. 279, p. 643-665. 1971: Geology and geochronology of the Hellroaring Creek Stock, British Columbia; Canadian Journal of Earth Sciences, v. 8, p. 85-95.

222 Thermobarometric Calculation of Peak Metamorphic Conditions

Schandl, E.S. and Davis, D.W. 2000: Geochronology of the Sullivan Deposit: U-Pb and Pb-Pb Ages of Zircons and Titanites; in The Geological Environment of the Sullivan Deposit, British Columbia, (ed.) J.W. Lydon, J.F. Slack, T. Höy, and M.E. Knapp; Geological Association of Canada, Mineral Deposits Division, MDD Special Volume No. 1, p. Schandl, E.S., Davis, D.W., and Gorton, M.P. 1993: The Pb-Pb age of metamorphic titanite in the chlorite-pyrite altered footwall of the Sullivan Zn-Pb SEDEX deposit, British Columbia, and its relationship to the ore; Geological Association of Canada/Mineralogical Association of Canada, Program with Abstracts, v. 18, p. A93. Scott, S.D. 1973: Experimental calibration of the sphalerite geobarometer; Economic Geology, v. 68, p. 466-474. Spear, F.S. 1991: On the interpretation of peak metamorphic temperature in light of garnet diffusion during cooling; Journal of Metamorphic Geology, v. 9, p. 379-388. Symmes, G.H. and Ferry, J.M. 1992: The effect of whole-rock MnO content on the stability of garnet in pelitic schists during metamorphism; Journal of Metamorphic Geology, v. 10, p. 221-237. Toulmin III, P., Barton, P.B. jr., and Wiggins, L.B. 1991: Commentary on the sphalerite geobarometer; American Mineralogist, v. 78 p. 1038-1051.

223